| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9
Comments:
<0> aedinius: i was using vmware server today... freakin' NICE
<0> i'm installing fedora core on my desktop to make use of it
<1> aedinius: why the yay?
<2> That's a VM on my MacBook
<1> lol
<2> Near native speeds.
<1> wow
<2> Only issue is the video -- parallels video is kinda slow.
<2> But not bad.
<1> what kinda processor?
<2> T2400
<1> dual core? ;/
<0> aedinius: main reason i like VMWare better? because minix's networking works under it
<0> ;p
<2> Vratha: Nice
<3> aedinius: what's up?
<2> wait wait
<4> someone still plays with minix
<2> something's wrong
<4> !
<0> jeffloc: yeah, a whole team over at a university in amsterdam just updated it to minix 3
<0> but i'm not really playing; i'm using it to do academic work
<4> is it bloating up like linux distros?
<2> Nice
<0> no, the base install ain't got ****
<2> http://igotallthe.info/~drew/scaleup.jpg
<0> but the kernel is now a true microkernel
<4> good
<3> jeffloc: How is linux distro bloated?
<4> by getting bigger and nigger and ,...
<4> eek!
<0> hahaha
<4> bigger not nigger
<4> !
<3> WHat part is getting bigger?
<4> I'd better look when I type
<4> all of it
<4> I'm pleased that nonongo wasn't here
<4> wouldn't that have set him off
<5> dodongo
<0> bobongo
<4> or just shorten it to nong
<6> hi
<6> what is your favorite PIC16Fxxx C compiler?
<7> mmm lucida typewritr font is better
<5> scul, #nop sometimes talks about these things
<5> i dont use the 16f.. only the 18f
<6> i'll try
<8> the one that comes with mplab ****s bad, at least on the 18f
<9> GCC has been ported, has it not?
<5> nope
<6> anyway, i'll use asm
<7> hi Aaron
<5> there is a gcc port to a related 16-bit target, but its not the same thing. that port isnt in official sources though.
<9> I thought I saw something about GCC and the 18f, maybe just vague plans that were never realized
<5> some people, including myself, have announced interest in working on a port, but noone at this time has implemented one.
<5> most of the people you see asking about ports have no idea what is even required to implement one :)
<6> 1. computer, 2. good will, 3. loads of 6-packs
<5> plus a few months, plus the ability to read documentation, plus the ability to actually learn about a large, real-life peice of software
<5> most people interested in gcc ports are students or similar, who have never done any of these, nor have any particular intention of doing so, once faced with the proposition.
<9> GCC sure is complex piece of software, I tried looking into it to improve the poor amd64 optimisation
<9> Or the poor register allocation in general
<6> http://sourceforge.net/projects/picgcc
<10> gcc is just an awful piece of software.
<5> scul, thats interesting
<5> once you get to know a particular part, generally you find it is exactly the way you would have written it, modulo various style issues that don't matter.
<9> Awful? No, but fairly large
<5> most of the interesting parts of gcc are fairly textbook implementations of the appropriate algorithms.
<9> Yes I'm not complaining about the way it's designed, but it is complex
<10> AaronWL: i write insane code, not normal code. i find that gcc is implemented in an incredibly goofy way.
<6> No File Packages Defined :)))
<6> it's fake project
<5> evilgeek, can you be more specific?
<5> scul, and the cvs is empty
<6> yes
<5> admin is David Santo Orcero .. who is this?
<10> the problem with gcc is that it's a huge project written in C. people tell me that it has some fairly serious maintainability issues.
<5> evilgeek, im getting the feeling you don't have a specific, first-hand criticism.
<9> A different language would not solve the maintainability issues of any huge software
<5> there is very strong sentiment to convert gcc to c++, mainly to catch additional catagories of bugs.
<5> and very strong resistance to this also, mainly by the people who don't know c++ :)
<5> in the gcc case, they want to switch to get a very specific set of features, which would catch a very specific and real catagory of bugs.
<10> AaronWL: not really. i haven't looked at gcc in depth. i had a vague desire to hack something into it at some point, and then i noticed that that would be hell.
<8> converting it to c++ would prolly create additional catagories of bugs
<9> A conversion to C++ sounds like a _very_ bad idea to me
<5> evilgeek, this is what i meant about the students and the three criteria students generally don't meet.
<10> Maloeran: various C++ features, if used judiciously, can boost maintainability.
<10> AaronWL: ?
<5> Maloeran, right, but based on my experience with similar people who say the same things, i tend to doubt the motives for your claims
<9> The C++ code I often see is less maintanable and less flexible ( and slower, and consumes more memory ) than typical good C code
<5> evilgeek, [04:19] <5> plus a few months, plus the ability to read documentation, plus the ability to actually learn about a large, real-life peice of software
<9> It's possible to write good C++, and that implies not relying on its fixed OO scheme when it is not appropriate
<5> Maloeran, and of course you're also willing to disclose you have limited experience working with c++ and a less than complete understanding of the language, right?
<5> 'OO' is irrelevent, because it means nothing.
<10> it's not ability in my case. it's laziness.
<9> Not quite, AaronWL, but I won't hide I prefer C
<10> Maloeran: yeah, but that's because there are dumb***es who write C++ code and give it a bad name.
<5> as i said, gcc wants to convert to c++ mainly to get a few specific c++ features that are generally agreed, by those who know, to fix specific problems gcc has.
<5> mainly, additional type-safety
<5> C has been shown to be inadequate in this area (virtually anything else modern would be better)
<8> c++ is often used in very bad ways that lead to hardcore obfuscated code
<10> AaronWL: i have very little desire to hack on a project that takes more than a few minutes to learn about.
<5> of the main people who want to do it, they're not adopting a significant paradigm change.. just introduction of a few additional features. so perhaps 'switch' is a bit dramatic.
<5> evilgeek, hehe, right
<9> What kind of specific problems are you refering to, AaronWL?
<10> does "you're trolling" now mean "i disagree"?
<5> Maloeran, lack of type safety leading to misuse of objects in inappropriate contexts
<9> You sure can get GCC to complain about type "misuse" if you so desire
<10> i always thought "trolling" referred to unenlightened comments crafted to inflame the rage of others.
<5> Maloeran, im talking about the internals of gcc, not its diagnostics.
<4> evilgeek, looked as if that was where he was going
<10> 'inflame' is clearly not the right word, but you get the idea.
<10> jeffloc: it does?
<5> the primary of concern, although there are many, are with the intermediate language-independent tree represenatation
<10> jeffloc: pointing out problems in a particular language does not constitute trolling.
<5> you tend to have tons of different kinds of nodes, that all need to conform to various hetereogenous interfaces, that are only allowed to be used in certain places, etc
<5> so what you want is to be able to take some peice, use it in a certain way, without knowing what it is, etc
<9> That could get very messy in C++ as well
<5> this is polymorphism, btw, and c++ does this very easily
<7> very_cautios_and_very_slow_dynamic_but_type_safe_cast<type123>(expression)
<7> that's c++ for you
<0> scientology++!!!
<10> the C++ casts piss me off.
<5> dynamic_cast is not within the proposed subset of c++ to be used.
<5> generally, if you just lay out your c++ types carefully (in a manner that is already done in gcc, just without effect), you get all of this automatic checking for free.
<5> c++ is all about automatic checking. half of the features of c++ over c are about telling the compiler a few rules you'd like to follow, then having the compiler check to make sure you follow them.
<5> its not just fluff.. its stuff intelligent, bad*** programmers consistantly screwed up with in C, and now is being caught automatically in C++
<5> im tired of hearing about the plea to 'just program smarter'
<5> most of us are probably already programming at 100% smartness
<6> most of C++ stuff kicks in during compile phase
<7> typical c++ programemrs can't write exception-safe copde
<5> (the other half of c++, thats not compile-time checking, is the ability to add abstraction without performance penalty)
<7> 100% of tehm can't
<5> 100% of typical programmers? :)
<0> whoa... there's a freakin' #scientology channel
<5> the easy way to see a programmer who doesn't have a clue how to write exception-safe code is the programmer who complains about how much auto_ptr ****s
<9> These abstractions can have a real cost, AaronWL, but C++ is often misused
<7> exceptions is worst feature of c++
<6> Aeon-_ - but it's easy: try{ /*your code*/} catch(...){/*hehe*/}
<5> Maloeran, C is often misused. asm is often misused. Aeon-_ misused.
<5> Aeon-_, i would say, they are the least intuitive, and worst taught, feature of c++
<6> Aeon_ - why? exceptions are optional. you do not have to use them.
<7> scul: wrong
<9> I tend to see C++ abused much more than C around me, it seems to let bad programmers do a worst job than they would in C
<5> but scul is right.. the beauty of c++ (opposed to java, for eg) is that you can only use the language features you see fit.
Return to
#c
or
Go to some related
logs:
esx sql
#politics
Tikachbila
#beginner
#networking
#c
mirc for ubuntu
#beginner
onyxia prequest
#mirc
|
|
